Pulmonary Deadspace and Postoperative
Outcomes in Neonates Undergoing Stage 1
Palliation Operation for Single Ventricle Heart
Disease*
Divya Shakti, MBBS, MPH1,2; Doff B. McElhinney, MD, MPH1,2; Kimberlee Gauvreau, ScD1,2;
Vamsi V. Yarlagadda, MD1,2; Peter C. Laussen, MBBS1,2; Peter Betit, RRT-NPS3; Mary L. Myrer, RN, RT3;
Ravi R. Thiagarajan, MBBS, MPH1,2
Objectives: Increased pulmonary dead space fraction (Vd/Vt) has
been associated with prolonged mechanical ventilation after surgery for congenital heart disease. The association of Vd/Vt with
clinical outcomes in neonates undergoing stage 1 palliation for
single ventricle congenital heart disease has not been reported.
We describe changes in Vd/Vt, differences in Vd/Vt based on
shunt type (right ventricle to pulmonary artery conduit vs modified
Blalock-Taussing shunt) and association of Vd/Vt with postoperative outcomes in patients undergoing stage 1 palliation.
Design: Retrospective chart review for demographic, hemodynamics, outcome information, and Vd/Vt values were collected at
6-hour intervals during the first 48 postoperative hours in neonates undergoing stage 1 palliation. Vd/Vt was calculated using
mixed expired Co2 (PeCo2) obtained from capnography and
paired arterial blood gas Co2 values.
Setting: Cardiac ICU in a tertiary care pediatric hospital.
Patients: Newborns with single ventricle congenital heart disease
undergoing stage 1 palliation during 2003–2004.
Measurements and Main Results: Of the 51 patients, 31 had right
ventricle to pulmonary artery and 20 had Blalock-Taussing shunt.
*See also p. 777.
1
Department of Cardiology, Boston Children’s Hospital, Boston, MA.
2
Department of Pediatrics, Harvard Medical School, Boston, MA.
3
Department of Respiratory Care, Boston Children’s Hospital, Boston, MA.
This study was performed at Boston Children’s Hospital, 300 Longwood
Ave, Boston, MA 02115.
Presented, in part, at the Pediatric Cardiac Intensive Care Society Annual
meeting, 2012.
Dr. Shakti received support for travel from the Boston Children's Heart
Foundation. Dr. Betit is employed by the Boston Children's Hospital. The
remaining authors have disclosed that they do not have any potential conflicts of interest.
For information regarding this article, E-mail: divya.shakti@cardio.
chboston.org
Copyright © 2014 by the Society of Critical Care Medicine and the World
Federation of Pediatric Intensive and Critical Care Societies
DOI: 10.1097/PCC.0000000000000226
728
www.pccmjournal.org
Although Vd/Vt was lower in the Blalock-Taussing shunt group over
all time points (p = 0.02), maximal Vd/Vt on day 1 (0.49 ± 0.07)
and on day 2 (0.46 ± 0.08) were not different between the shunt
groups. Vd/Vt decreased significantly over time in both shunt
groups (p = 0.001 for right ventricle to pulmonary artery; p < 0.001
for Blalock-Taussing shunt). Higher maximal Vd/Vt during first 48
postoperative hours was independently associated with fewer ventilator (β = −26.6; p = 0.035) and hospital-free days in the first
month after stage 1 palliation (β = −40.4; p = 0.002) after adjusting
for potential confounders in a multivariable linear regression model.
Conclusions: Increased pulmonary dead space exists early after
stage 1 palliation operation for single ventricle congenital heart
disease. Higher Vd/Vt during the first 48 postoperative hours was
associated with longer duration of ventilation and hospital LOS
and may be a useful marker of postoperative outcomes in this
population. (Pediatr Crit Care Med 2014; 15:728–734)
Key Words: hypoplastic left heart syndrome; Norwood operation;
postoperative outcomes; pulmonary dead space ventilation;
ventilation-perfusion abnormality
P
ulmonary dead space is the portion of inhaled air in
the alveoli that does not participate in gas exchange
and quantifies the magnitude of ventilation-perfusion
abnormalities in the lung. Pulmonary dead space fraction can
be easily measured in patients receiving mechanical ventilation
using bedside volumetric capnometry and PaCO2 values (1).
The reported pulmonary deadspace fraction (Vd/Vt)
in normal children ranges from 0.25 to 0.33 (2–4). Vd/Vt is
increased in parenchymal lung disease and in diseases that alter
lung perfusion. High Vd/Vt is associated with increased mortality in children with congenital diaphragmatic hernia and in
adults and children with ALI (5–8). Children with congenital
heart disease (CHD) characterized by either pulmonary hypoperfusion or cyanosis have been shown to have higher Vd/Vt,
October 2014 • Volume 15 • Number 8
Cardiac Intensive Care
and higher Vd/Vt has been shown to prolong duration of
mechanical ventilation and length of hospitalization after cardiac surgery for CHD (9, 10). More recently, an association
between larger alveolar deadspace fraction using end-tidal Co2
(Etco2) and increased mortality was noted in pediatric patients
with acute hypoxemic respiratory failure (11).
In patients undergoing stage 1 palliation (S1P), increased
Vd/Vt may occur in the presence of reduced output from the
systemic ventricle, obstruction to pulmonary blood flow, or the
presence of lung disease. Higher Vd/Vt in this population may
result in the need for prolonged mechanical ventilation and
poorer clinical outcomes; however, these have not been previously described. Thus, our goals were to characterize changes
in Vd/Vt during the first 48 postoperative hours in neonates
undergoing S1P, to investigate whether Vd/Vt varied by shunt
type (right ventricle to pulmonary artery conduit [RV-PA] vs
modified Blalock-Taussing shunt [BTS]), and to explore the
association of Vd/Vt with postoperative outcomes in patients
with single ventricle CHD undergoing S1P operation.
MATERIALS AND METHODS
The Committee for Clinical Research at Boston Children’s
Hospital approved data collection for purposes of this retrospective cohort study, and the need for informed consent was
waived. We included neonates undergoing S1P with either
BTS or RV-PA during January 1 2003, to June 30, 2004, and
managed postoperatively in the cardiac ICU (CICU) at Boston Children’s Hospital. This study period was chosen because
some of the early postoperative variables needed for the study
were collected as a part of a quality improvement initiative to
assess the use of Etco2 monitoring in this group of patients. We
excluded patients who died in the operating room or within 48
hours of CICU admission, patients placed on extracorporeal
membrane oxygenation (ECMO) intraoperatively or within 48
hours of CICU admission, those extubated within 48 postoperative hours, and patients with significant endotracheal tube
leak (> 30% reduction in exhaled tidal volume compared with
inspired tidal volume). During this time period, S1P operation
with either a BTS or a RV-PA was performed according to the
discretion of the patient’s surgeon and cardiologist. Postoperatively, all patients were managed with pressure controlled synchronized intermittent mandatory ventilation aiming for tidal
volume of 10–12 mL/kg, per unit policy.
Data Collection
Patients’ medical records were reviewed to collect data on
demographics, diagnosis, perioperative information, and study
outcome variables. Data on Etco2, mixed expired Co2 (PeCo2),
and corresponding PaCO2 values from arterial blood gas measurements during mechanical ventilation were obtained from
a database maintained by the Respiratory Care Department.
The database contained these values collected every 6 hours
for the first 48 postoperative hours and at the time of extubation. During this time period, real-time single breath PeCo2 and
etco2 were monitored with CO2SMO-Plus noninvasive capnograph monitor (Novametrix Medical Systems, Wallingford,
Pediatric Critical Care Medicine
CT) using a single sensor integrated with a pneumotachometer
measuring exhaled Co2, air flow, and pressure. This sensor was
placed between the endotracheal tube and ventilator circuit
(CAPNOGARD monitor and CAPNOSTAT Mainstream CO2
Sensor, Respironics Pulse Oximeter and CO2 monitor; Philips
Respironics, Andover, MA). For calculation of Vd/Vt, simultaneously obtained PeCo2 values and PaCO2 values were used.
Physiologic Vd/Vt was calculated using the Enghoff modification of Bohr’s equation, Vd/Vt = (PaCO2 – PeCo2)/PaCO2 (1).
Vital signs and ventilator settings were recorded at the same
time from chart review. Decisions regarding patient management, including readiness for extubation, were guided by conventional parameters and were made by physicians caring for
the patient. Vd/Vt data were not available to the clinicians and
were not used to guide weaning from ventilation.
Postoperative inotrope use was quantified using an inotrope
score developed by Wernovsky et al (12). Postoperative outcomes
included length of mechanical ventilation, length of CICU and
hospital stay (discharge home or transfer to outside facility), and
extubation outcome. The outcomes of number of ventilator-free
days and hospital-free days were defined as the number of days
the patient did not need mechanical ventilation and was not hospitalized, respectively, within 30 days from the date of S1P.
Statistics
We compared physiological parameters between the two shunt
groups using the t test, Wilcoxon rank-sum test, or Fisher exact
test as appropriate. Changes in Vd/Vt and hemodynamic parameters over time were analyzed using linear mixed-effects modeling. Associations between Vd/Vt and postoperative outcomes
were assessed using linear regression and survival analysis. The
elements chosen for adjustment in the multivariable models
were those that showed bivariate association (p < 0.1) with the
outcome in question. These variables were analyzed in a multivariable regression model with backward elimination technique
and final model included covariates that were significant at a p
value of less than 0.05. Statistical analyses were conducted using
STATA version 10 (StataCorp LP, College Station, TX) and SAS
version 9.3 (SAS Institute, Cary, NC). Statistical significance was
set at a p value of less than or equal to 0.05. Continuous variables
are summarized as either mean ± standard or median (IQR),
and categorical variables as frequency and percent.
RESULTS
Study Population
During the study period, 66 patients underwent S1P. Of these,
15 patients were excluded (Vd/Vt data were not available in 11,
three were placed on ECMO within 48 hours of surgery, and
one died in the operating room). Demographic, preoperative,
and intraoperative data on 51 patients with single ventricle
CHD (mostly hypoplastic left heart syndrome) who underwent S1P (RV-PA = 31 and BTS = 20) during the study period
and had Vd/Vt data are shown in Table 1. For the entire cohort,
median gestational age was 37 weeks (range, 35–38 wk). Three
patients in the RV-PA group died before hospital discharge.
www.pccmjournal.org
729
Shakti et al
Table 1.
Demographic, Preoperative, and Intraoperative Variables for the Study Cohort
Right Ventricle to Pulmonary
Artery Conduit (n = 31)
Variable
Gender (women) (%)
8 (25.8)
Birth weight (kg)
3.2 (2.4–3.4)
Blalock-Taussing Shunt
(n = 20)
p
12 (60)
0.02
3.3 (2.6–3.4)
0.52
Birth weight < 2.5 kg (%)
9 (29)
4 (20)
0.53
Age at surgery (d)
6 (5–8)
5 (4–10.5)
0.62
Genetic syndrome (%)
1 (3.2)
1 (5)
1
Noncardiac anomaly (%)
3 (9.7)
1 (5)
0.37
Mod-severe atrioventricular valve regurgitation
1 (3.2)
1 (5)
1
Ventricular dysfunction
2 (6.4)
1 (5)
1
Emergency atrial septum intervention
2 (6.4)
1 (5)
1
Cardiac risk factors (%)
Cardiac anatomy (%)
0.74
HLHS-mitral atresia/AA
9 (29)
6 (30)
HLHS-MS/aortic stenosis
9 (29)
7 (35)
HLHS-MS/AA
9 (29)
3 (15)
HLHS variant
2 (6.4)
1 (5)
Other
2 (6.4)
3 (15)
Intraoperative
Total pump time (min)
140 (113–163)
Cross clamp time (min)
Shunt size (%)
60 (51–83)
5 mm: 28 (90.3); 4 mm: 3 (9.7)
150 (142.5–172.5)
0.03
53.5 (47.5–65.5)
0.19
3.5 mm: 20 (100)
HLHS = hypoplastic left heart syndrome; AA = aortic atresia, MS = mitral stenosis, HLHS variant = mitral atresia/MS, aortic stenosis, ventricular septal defect.
Data are presented as n (%) or median (interquartile range). Boldface values indicate statistically significant comparisons.
Postoperative Hemodynamic, Respiratory, and
Outcome Variables
Changes in hemodynamics and oxygenation parameters during the first 48 hours postoperatively for the two groups are
depicted in Figure 1, A–D. As expected, higher diastolic pressures were noted in RV-PA group (Fig. 1C). Pao2/Fio2 ratio
increased significantly over time for both groups but remained
significantly lower in the RV-PA group (Fig. 1D). The postoperative indices of tissue perfusion, inotrope support, and postoperative outcomes were similar for both groups as shown in
Table 2. The tidal volumes and pulmonary compliance did not
show any significant change over time and were not different
between the shunt groups.
Postoperative Vd/Vt
Data on postoperative Vd/Vt values over time for the all
patients and stratification by the shunt type during the study
period are shown in Figure 2A. Vd/Vt significantly decreased
over time for both shunt types. The BTS group had lower
Vd/Vt values when compared with RV-PA group at all time
points (p = 0.02). Table 3 shows the mean Vd/Vt values on
days 1 and 2 and maximum Vd/Vt during the study period. As
730
www.pccmjournal.org
shown in Table 3, the maximum Vd/Vt was higher for RV-PA,
but the mean Vd/Vt on postoperative days 1 and 2 was similar
for the two shunt groups. Figure 2B shows arterial to end-tidal
Co2 gradient (AEG) during the first 48 postoperative hours
and that at the time of extubation for all patients and stratified
by the shunt type. Vd/Vt and AEG values were highly correlated at all time points.
Association With Clinical Outcomes
Table 4 shows bivariate associations, of mean Vd/Vt on days 1
and 2 and maximum Vd/Vt during the first 48 postoperative
hours and postoperative outcomes. Here, a higher maximum
Vd/Vt was associated with fewer (negative β coefficient) number of ventilator-free days and hospital-free days and increase
(positive β coefficient) in the log length of mechanical ventilation and length of hospital stay.
Table 5 shows associations between Vd/Vt and the postoperative outcomes if a significant bivariate association was noted,
after adjusting for confounders (multivariable models specified
in the footnotes of Table 5). After adjusting for confounders,
higher maximum Vd/Vt during the first 48 postoperative hours
was independently associated with fewer ventilator-free days.
October 2014 • Volume 15 • Number 8
Cardiac Intensive Care
Figure 1. Hemodynamics and oxygenation parameters during the first 48 postoperative hours after stage 1 palliation operation for single ventricle
congenital heart disease. A, Heart rate; B, systolic blood pressure; C, diastolic blood pressure; D, Pao2/Fio2 ratio. Data shown stratified by shunt type
used to provide pulmonary blood flow. A value of p for each shunt group, representing trend over time for each group obtained by linear mixed models,
is shown. The p value comparing overall differences between the two groups is shown on the bottom left of the figures. BTS = Blalock-Taussing shunt,
RV-PA = right ventricle to pulmonary artery conduit.
Both higher mean Vd/Vt on day 1 and maximum Vd/Vt were
independently associated with fewer hospital-free days. We used
maximal Vd/Vt values to categorize the cohort into two groups
Table 2.
based on a Vd/Vt value of less than or equal to 0.5 and greater than
0.5 and evaluated association with clinical outcomes (10, 13).
Maximal Vd/Vt greater than 0.5 was noted in 43 newborns
Postoperative Variables and Outcomes
Variable
Maximum post cardiopulmonary bypass
lactate (mmol/L)
Right Ventricle to Pulmonary
Artery Conduit (n = 31)
Blalock-Taussing
Shunt (n = 20)
p
7.15 (5.8–8.5)
6.3 (5.8–8.5)
0.61
Average lactate in first 12 hr
2.1 (1.6–4.5)
2.5 (2–5.1)
0.33
Ionotropic score
10 (5–12.5)
7.5 (5–10)
0.48
Time to extubation (d)
5 (4–9.5)
Duration of ICU stay (d)
7 (5.5–13)
6 (4.5–8.5)
0.6
9.5 (7–18.5)
0.15
Duration of hospital stay (d)
16 (12–26)
19 (11–128)
0.65
Ventilator-free days
25 (14–26)
24 (20.5–25)
0.92
11 (0–19)
0.62
Hospital-free days
11.5 (0–18)
Data are presented as median (interquartile range).
Pediatric Critical Care Medicine
www.pccmjournal.org
731
Shakti et al
Figure 2. A, Postoperative physiologic deadspace ratio (Vd/Vt) during the first 48 postoperative hours for the entire cohort and for the two shunt
types. B, Arterial to end-tidal Co2 gradient (AEG) during the first 48 postoperative hours and at the time of extubation for the entire cohort and for the
two shunt types. A value of p for each shunt group, representing trend over time for each group obtained by linear mixed models, is shown. The p value
comparing overall differences between the two groups is shown on the bottom left of the figures. BTS = Blalock-Taussing shunt, RV-PA = right ventricle
to pulmonary artery conduit.
(84%), more frequently in RV-PA than in BTS (29 [94%] vs 14
[70%]; p = 0.04). Patients with maximal Vd/Vt greater than 0.5
had significantly fewer number of ventilator-free days when
compared with those with Vd/Vt less than or equal to 0.5 (19.7
vs 23.5; p = 0.04). However, the duration of ventilation (log rank
p = 0.34) and length of hospital stay (log rank p = 0.84) based on
this categorization were not different between the two groups.
There were 12 (24%) extubation failures in the entire cohort.
Proportion of patients failing extubation was not significantly
different in the two groups (10 [32%] vs 2 [10%] in RV-PA vs
BTS; p = 0.06). When looking at the entire cohort, there was a
trend toward higher maximum Vd/Vt (0.61 vs 0.56; p = 0.05)
and higher maximum AEG at extubation (16.8 vs 12.3; p = 0.08)
in those failing extubation when compared with those extubated
successfully. Maximum Vd/Vt and AEG correlated significantly
regardless of outcome of extubation. There were 10 patients
(19.6%) in the entire cohort in whom Vd/Vt failed to decrease
from 6 to 48 hours. Extubation failure was higher in patients
showing no change or increase in Vd/Vt when compared with
those with decrease in Vd/Vt during the first 48 postoperative
hours (5 [50%] vs 34 [17%]; p = 0.04). Four of five were in
RV-PA group. Similar findings were noted using change in AEG
during the 48 postoperative hours and extubation failure.
DISCUSSION
In our cohort of neonates with single ventricle CHD undergoing S1P, we found that large pulmonary dead space (Vd/Vt)
exists in the early postoperative period. Vd/Vt was lower in the
BTS group when compared with that in the RV-PA group at all
time points. Vd/Vt decreased significantly in both shunt types.
Although Vd/Vt values decreased over the early postoperative period, a large difference in arterial to end-tidal CO2 (thus
Vd/Vt) remained at the time of extubation from mechanical
ventilation. We found that higher Vd/Vt was associated with
the prolonged duration of postoperative mechanical ventilation and length of hospital stay. These findings suggest that
significant deadspace ventilation is present in the early postoperative period following S1P and that the magnitude of
pulmonary dead space may be associated with postoperative
outcomes in these patients.
A significant finding in our study is the association of
Vd/Vt with duration of postoperative ventilation. Although
we do not know the exact etiology of increased Vd/Vt in our
cohort, some possible explanations include pulmonary hypoperfusion from postoperative low-cardiac output syndrome
commonly seen in the first 24 postoperative hours after neonatal cardiac surgery (12). Alternatively, the large Vd/Vt may
Pulmonary Deadspace Fraction (Vd/Vt) Summary During the First 48
Postoperative Hours
Table 3.
All Patients (n = 51)
Right Ventricle to Pulmonary
Artery Conduit (n = 31)
Blalock-Taussing
Shunt (n = 20)
p
Maximum
0.57 (0.08)
0.59 (0.06)
0.54 (0.08)
0.02
Mean on day 1
0.49 (0.07)
0.50 (0.08)
0.47 (0.06)
0.06
Mean on day 2
0.46 (0.08)
0.48 (0.08)
0.44 (0.07)
0.09
Vd/Vt at Time (hr)
Data are presented as mean (sd). Boldface value indicate statistically significant comparisons.
732
www.pccmjournal.org
October 2014 • Volume 15 • Number 8
Cardiac Intensive Care
Bivariate Associations Between Pulmonary Deadspace Fraction (Vd/Vt) and
Postoperative Outcomes
Table 4.
Log Length of
Mechanical Ventilation
Mean Vd/Vt on day 1
1.4 (0.30)
Mean Vd/Vt on day 2
–0.32 (0.79)
Maximum Vd/Vt
Ventilator-Free
Days
–27.4 (0.095)
1.15 (0.94)
2.2 (0.096)
Log Length of ICU
Stay
1.3 (0.39)
–0.36 (0.78)
–43.1 (0.006)
1.9 (0.18)
Log Length of
Hospital Stay
Hospital-Free
Days
3.8 (0.013)
–52.9 (0.001)
1.4 (0.30)
3.7 (0.011)
–23 (0.12)
–48.3 (0.002)
Data presented as β coefficients (p value). Boldface values indicate statistically significant comparisons.
have resulted from postoperative management after the S1P
that aims at limiting pulmonary blood flow in favor of systemic perfusion, or anatomical issues related to obstruction of
the shunt providing pulmonary blood flow. We do not have
concomitant imaging correlates, such as echocardiogram to
support a hypothesis of possible obstruction to shunt flow.
Although the exact etiology of Vd/Vt in our cohort could not
be determined, we found poorer clinical outcomes in patients
with increased Vd/Vt during the first 48 postoperative hours.
The presence of high AEG at the time of extubation suggests
that Vd/Vt abnormalities may persist in these patients even
after the immediate post-operative period. Regardless, high Vd/
Vt has significant clinical implications. Ventilatory inefficiency
because of large Vd/Vt results in the need for a higher minute
ventilation to exhale a given Co2 load. This is often achieved
with increased respiratory rate. Tachypnea as a compensation
of high Vd/Vt during weaning can delay readiness for extubation resulting in longer duration of ventilation. Tachypnea and
increased respiratory effort may contribute to increased caloric
need, and poor growth, which is often seen in these patients.
Although postextubation Vd/Vt values were not measured as
part of this study, persistence of large Vd/Vt postextubation
may result in persistent tachypnea even after extubation and
can delay oral feeding resulting in longer hospitalization. Our
finding that higher Vd/Vt is associated with prolonged ventilation is similar to those described by Fletcher et al (9) and Ong
et al (10) in postoperative patients undergoing repair of CHD.
Ong et al (10) measured Vd/Vt during the first 24 postoperative hours and found higher Vd/Vt in patients with residual intra cardiac shunt physiology than those without shunt
physiology (0.5 vs 0.4). Fletcher et al (9) reported high Vd/Vt
(0.41) in patients with pulmonary hypoperfusion. These studies were conducted in a more heterogeneous population based
on age and type of CHD (9, 10), indicating that association
of high Vd/Vt with prolonged duration of mechanical ventilation may also be seen in patients with other forms of CHD and
those undergoing reparative or palliative surgical procedures.
Hubble et al (13) in a study of patients with lung disease
receiving mechanical ventilation found that Vd/Vt less than
or equal to 0.5 was a reliable predictor of extubation success and Vd/Vt greater than 0.65 predicted extubation failure. There were very few patients in our cohort with Vd/Vt
greater than 0.65 precluding evaluation for its association with
extubation outcomes. We only found a trend toward higher
maximum Vd/Vt and higher AEG in patients failing extubation. Although we were unable to define a cutoff point for
Vd/Vt values that predicted extubation failure, we found that a
maximum Vd/Vt greater than 0.5 was significantly associated
with prolonged duration of postoperative mechanical ventilation. This may suggest that postoperative S1P patients with
Vd/Vt greater than 0.5 may require careful evaluation for
potentially treatable causes of pulmonary hypoperfusion, such
as anatomical obstructions to the shunt or pulmonary arteries
to help reduce duration of mechanical ventilation in this population. Furthermore, because capnography and arterial access
for blood gas monitoring are almost universal in this patient
population postoperatively, Vd/Vt can be easily estimated and
could serve as a marker to help identify those at higher risk of
needing prolonged mechanical ventilation in the postoperative
period after S1P. Riou et al (14) have shown that higher Vd/Vt
Multivariable Associations Between Pulmonary Deadspace Fraction (Vd/Vt) and
Postoperative Outcomes
Table 5.
Log Length of Mechanical
Ventilationa
Mean Vd/Vt on day 1
Maximum Vd/Vt
Ventilator-Free
Daysb
Log Length of
Hospital Stayc
Hospital-Free
Daysd
—
–7.5 (0.572)
1 (0.46)
–40.3 (0.004)
1.5 (0.20)
–26.6 (0.035)
1.9 (0.13)
–40.4 (0.002)
Models adjusted for the following:
a
Birth weight < 2.5 kg and average heart rate on postoperative day 1.
b
Noncardiac anomaly, birth weight < 2.5 kg, prenatal intervention, and cardiopulmonary by-pass time.
c
Noncardiac anomaly, emergent preoperative atrial septum intervention, average heart rate on postoperative day 1, and average Pao2/Fio2 ratio on postoperative day 1.
d
Noncardiac anomaly, emergent preoperative atrial septum intervention, and average heart rate on postoperative day 1.
Data presented as β coeffecient (p value). Boldface values indicate statistically significant comparisons. Dash indicates that multivariable analysis for log length
of mechanical ventilation and mean Vd/Vt on day 1 was not performed since there was no significant bivariate association noted.
Pediatric Critical Care Medicine
www.pccmjournal.org
733
Shakti et al
value at extubation predicted need for noninvasive ventilation
and prevented reintubation in children receiving mechanical
ventilation for respiratory failure. We did not evaluate the use
of noninvasive ventilation in our cohort. It is possible that the
patients with higher Vd/Vt in our cohort may have benefitted
from aggressive use of noninvasive ventilation.
We showed that Vd/Vt values and AEG values were higher
for neonates undergoing S1P with RV-PA when compared with
those undergoing S1P with BTS. The RV-PA shunt group had
a higher proportion of patients with maximum Vd/Vt values
greater than 0.5. In absence of any differences in tidal volumes
and compliance between the shunt types, these findings may
suggest more restricted pulmonary blood in the RV-PA conduit
group. This is consistent with our finding of significantly lower
Pao2/Fio2 ratio in the RV-PA group when compared with that in
the BTS group (15). Although the RV-PA group had larger shunts
when compared with the BTS group (5 mm RV-PA vs 3.5 mm
BTS), free regurgitation from the valve-less conduit during diastole can reduce net pulmonary blood flow resulting in increased
ventilator dead space.
Several limitations of our data need to be considered when
interpreting results from these analyses. First, our results cannot be generalized to patients who met our exclusion criteria.
The retrospective nature of this study makes it prone to
selection bias and effect of confounders. Our findings of prolonged ventilation and duration of hospitalization in patients
with higher Vd/Vt may be subject to factors not captured in
our dataset. We used an older cohort of patients in our analysis. Although there has been refinement in the operative techniques and improved outcomes in neonates undergoing S1P, we
feel that the physiology and ventilator management for these
patients have not changed significantly during this time period
in our institution. Similarly, the equipment to assess Vd/Vt may
have changed and refined over time; however, the principle of
measurement of Vd/Vt is the same. We could not obtain a baseline value of Vd/Vt before surgery, at extubation, and postextubation from mechanical ventilation to confirm that high Vd/Vt
occurred after S1P and that it persisted after the first 48 postoperative hours. Data on the contribution of the CO2SMO device
to dead space were not collected. However, the same device was
used for all patients. The small cohort of patients limited our
ability to detect smaller differences, thus allowing only larger differences to be identified statistically. Because of these limitations,
the clinical usefulness of Vd/Vt in identifying patients at risk for
need for prolonged mechanical ventilation after S1P should be
prospectively studied in a larger cohort of patients in the future.
CONCLUSIONS
We found high pulmonary dead space in the early postoperative period in neonates with single ventricle CHD undergoing
S1P. Higher Vd/Vt was noted in neonates with RV-PA when
compared with those with BTS group. Higher Vd/Vt was
734
www.pccmjournal.org
associated with poorer clinical outcomes, such as longer periods of mechanical ventilation, extubation failure, and longer
hospital stay. Increased Vd/Vt may result in impaired respiratory efficiency, causing higher respiratory rates and metabolic demand that may contribute to the need for prolonged
mechanical ventilation and hospitalization in these patients.
Pulmonary deadspace monitoring may serve as a useful noninvasive adjunct to the postoperative monitoring of single ventricle CHD patients undergoing S1P. Because Vd/Vt and AEG
were highly correlated, AEG could be used as a surrogate of
Vd/Vt.
REFERENCES
1. West JB: Ventilation. In: Respiratory Physiology - The Essentials.
Fifth Edition. Coryell PA (Ed). Baltimore, Williams & Wilkins, 1995, pp
11–20
2. Kerr AA: Dead space ventilation in normal children and children with
obstructive airways diease. Thorax 1976; 31:63–69
3. Nelson NM, Prodhom LS, Cherry RB, et al: Pulmonary function in
the newborn infant. V. Trapped gas in the normal infant’s lung. J Clin
Invest 1963; 42:1850–1857
4. Cook CD, Cherry RB, O’Brien D, et al: Studies of respiratory physiology in the newborn infant. I. Observations on normal premature and
full-term infants. J Clin Invest 1955; 34:975–982
5. Arnold JH, Bower LK, Thompson JE: Respiratory deadspace measurements in neonates with congenital diaphragmatic hernia. Crit
Care Med 1995; 23:371–375
6. Cepkova M, Kapur V, Ren X, et al: Pulmonary dead space fraction
and pulmonary artery systolic pressure as early predictors of clinical
outcome in acute lung injury. Chest 2007; 132:836–842
7. Nuckton TJ, Alonso JA, Kallet RH, et al: Pulmonary dead-space fraction as a risk factor for death in the acute respiratory distress syndrome. N Engl J Med 2002; 346:1281–1286
8. Coss-Bu JA, Walding DL, David YB, et al: Dead space ventilation in
critically ill children with lung injury. Chest 2003; 123:2050–2056
9. Fletcher R, Niklason L, Drefeldt B: Gas exchange during controlled
ventilation in children with normal and abnormal pulmonary circulation: a study using the single breath test for carbon dioxide. Anesth
Analg 1986; 65:645–652
10. Ong T, Stuart-Killion RB, Daniel BM, et al: Higher pulmonary dead
space may predict prolonged mechanical ventilation after cardiac surgery. Pediatr Pulmonol 2009; 44:457–463
11. Ghuman AK, Newth CJ, Khemani RG: The association between
the end tidal alveolar dead space fraction and mortality in pediatric
acute hypoxemic respiratory failure. Pediatr Crit Care Med 2012; 13:
11–15
12. Wernovsky G, Wypij D, Jonas RA, et al: Postoperative course and
hemodynamic profile after the arterial switch operation in neonates
and infants. A comparison of low-flow cardiopulmonary bypass and
circulatory arrest. Circulation 1995; 92:2226–2235
13. Hubble CL, Gentile MA, Tripp DS, et al: Deadspace to tidal volume
ratio predicts successful extubation in infants and children. Crit Care
Med 2000; 28:2034–2040
14. Riou Y, Chaari W, Leteurtre S, et al: Predictive value of the physiological deadspace/tidal volume ratio in the weaning process of mechanical ventilation in children. J Pediatr (Rio J) 2012; 88:217–221
15. Cua CL, Thiagarajan RR, Gauvreau K, et al: Early postoperative
outcomes in a series of infants with hypoplastic left heart syndrome
undergoing stage I palliation operation with either modified BlalockTaussig shunt or right ventricle to pulmonary artery conduit. Pediatr
Crit Care Med 2006; 7:238–244
October 2014 • Volume 15 • Number 8